Blue Light, Melatonin and Blue Light Blockers: What the 2026 Evidence Actually Supports

The most clinically useful blue light paper of 2026 so far is not a trial at all. Terán and colleagues, publishing in Scientific Reports in January, characterised the spectral emissions of 52 individual lamps across LED, incandescent and compact fluorescent technologies and computed melanopic illuminance and an estimated melatonin suppression value for each.^[1]

The spread is large enough to be actionable. Cool-white LED (median 12.3% MSV) and cool-white CFL (12.1%) suppressed estimated melatonin roughly three to eight times more than warm-white LED (3.6%), warm-white CFL (2.6%) or incandescent (1.5%) sources. Four tunable-CCT LED lamps were the standout intervention: dropping the same fixture from 5700 K to 2100 K reduced estimated suppression from 10% to 0.1% — a ~100-fold change achieved by turning a dial rather than buying eyewear.

The counterweight is the November 2025 systematic review and meta-analysis by Luna-Rangel and colleagues in Frontiers in Neurology, which pooled the only three double-blind crossover RCTs that used actigraphy as the outcome (total n = 49).^[2] Every pooled estimate crossed zero:

Glickman and colleagues addressed this directly by proposing melanopic daylight filtering density (mDFD) — a single, spectrally weighted number describing how much melanopic input a filter actually removes — and applying it to 26 commercially available pairs of glasses.^[3] Products varied substantially, and only those reaching mDFD ≥ 1 delivered reductions large enough to justify the claim on the box.

The Terán lens data point the same way. Of eight blue-light-filtering lenses tested, six offered only modest benefit relative to a standard clear lens. The two that reduced estimated melatonin suppression below 0.3% were both distinguished by a visible brown tint.^[1] This is the practical translation of an unglamorous optical fact: a lens that removes enough short-wavelength energy to change melanopic signalling cannot look clear. Clear "computer glasses" and amber or red evening lenses are not the same product category, and pooling them is part of why the trial literature looks flat.

Montanari and colleagues, publishing in JMIR mHealth and uHealth in June 2026, instrumented 21 adults for seven consecutive days with wearable EEG for sleep staging, HOBO loggers for ambient light, accelerometry, and an app logging smartphone use.^[4] Mixed-effects models estimated day-to-day associations with the following night's sleep.

Three findings stand out. Each additional hour of daytime exposure above 1000 lux predicted +10.67 minutes of total sleep time (95% CI 0.6 to 20.7) the following night, plus a 0.3-percentage-point reduction in N1 light sleep. Each minute of pre-bed smartphone use predicted +0.2 minutes of sleep onset latency (95% CI 0.0 to 0.4) — roughly 6 minutes per half-hour of scrolling. And measured ambient artificial light in the two hours before sleep onset showed no independent association with sleep outcomes.

With n = 21 this is a small, observational, hypothesis-generating study and should be read as preliminary. But it aligns with a pattern now visible across the field: daytime light exposure is an underused lever, and the phone in the hand may be doing more through arousal, content and time displacement than through photons.

In a five-week crossover trial, 39 myopic Japanese boys aged 10–12 wore either a partial blue-blocking lens (40% cut) or a matched clear corrective lens for the three hours before habitual bedtime, in two-week blocks separated by a one-week washout.^[6] Salivary melatonin did not differ between conditions. Yet the sleep phase advanced significantly — bedtime 22.03 ± 0.08 h versus 22.13 ± 0.09 h (p = 0.040) and sleep onset 22.26 ± 0.08 h versus 22.36 ± 0.10 h (p = 0.041) — with effects stronger in the second week and accompanied by reduced daytime irritability and disruptive behaviour and improved morning mood.

The magnitude is small (roughly six minutes) and the population narrow. Its interest is mechanistic: a dissociation between melatonin secretion and behavioural sleep timing suggests filters may act partly through alertness and arousal pathways that salivary melatonin and actigraphy both fail to capture — the same possibility raised in the 2025 meta-analysis discussion.^[2]

Campo-Beamud and Roig-Ferreruela's narrative review in Archivos de la Sociedad Española de Oftalmología, available online 20 July 2026, surveyed the 2009–2025 literature and concluded that blue-filtering lenses do not produce clinically meaningful improvements in visual acuity or digital eye strain symptoms and show small, inconsistent effects on sleep outcomes; that under normal viewing conditions screen light sits well below thresholds associated with photochemical retinal damage in experimental models; and that current evidence does not support routine preventive or therapeutic prescription of blue-blocking glasses.^[5]

That conclusion is defensible and should be repeated to patients who arrive expecting eyewear to protect their retinas or cure eye strain — it will not. Their recommended alternative is also the one the physiology supports: light hygiene, reduced nocturnal exposure, and patient education.